diff mbox

USB: ohci-omap: defer probe if PHY is missing

Message ID 20170102205355.18920-1-aaro.koskinen@iki.fi (mailing list archive)
State New, archived
Headers show

Commit Message

Aaro Koskinen Jan. 2, 2017, 8:53 p.m. UTC
Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
to be loaded to get the PHY going and ohci-omap should wait for those.

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
---
 drivers/usb/host/ohci-omap.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

Comments

Greg Kroah-Hartman Jan. 3, 2017, 4:51 p.m. UTC | #1
On Mon, Jan 02, 2017 at 10:53:55PM +0200, Aaro Koskinen wrote:
> Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
> to be loaded to get the PHY going and ohci-omap should wait for those.
> 
> Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>

Is this a new bug?  The 770 has been around for forever, why has this
not been a problem before now?

thanks,

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Aaro Koskinen Jan. 3, 2017, 6:50 p.m. UTC | #2
Hi,

On Tue, Jan 03, 2017 at 05:51:43PM +0100, Greg Kroah-Hartman wrote:
> On Mon, Jan 02, 2017 at 10:53:55PM +0200, Aaro Koskinen wrote:
> > Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
> > to be loaded to get the PHY going and ohci-omap should wait for those.
> > 
> > Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> 
> Is this a new bug?  The 770 has been around for forever, why has this
> not been a problem before now?

PHY/OTG support for 770 has been around only since v3.14. I was previously
compiling phy-tahvo as built-in, and bumped into this issue when compiling
it as a module.

A.
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Greg Kroah-Hartman Jan. 3, 2017, 7:05 p.m. UTC | #3
On Tue, Jan 03, 2017 at 08:50:31PM +0200, Aaro Koskinen wrote:
> Hi,
> 
> On Tue, Jan 03, 2017 at 05:51:43PM +0100, Greg Kroah-Hartman wrote:
> > On Mon, Jan 02, 2017 at 10:53:55PM +0200, Aaro Koskinen wrote:
> > > Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
> > > to be loaded to get the PHY going and ohci-omap should wait for those.
> > > 
> > > Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> > 
> > Is this a new bug?  The 770 has been around for forever, why has this
> > not been a problem before now?
> 
> PHY/OTG support for 770 has been around only since v3.14.

3.14 was released March, 2014.  A long time ago in kernel development :)

> I was previously compiling phy-tahvo as built-in, and bumped into this
> issue when compiling it as a module.

Why not just make the defconfig as built-in?

I'm not objecting to this patch, just really curious why no one else has
ever hit it.

thanks,

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Felipe Balbi Jan. 3, 2017, 7:11 p.m. UTC | #4
Hi,

Greg Kroah-Hartman <gregkh@linuxfoundation.org> writes:
> On Tue, Jan 03, 2017 at 08:50:31PM +0200, Aaro Koskinen wrote:
>> Hi,
>> 
>> On Tue, Jan 03, 2017 at 05:51:43PM +0100, Greg Kroah-Hartman wrote:
>> > On Mon, Jan 02, 2017 at 10:53:55PM +0200, Aaro Koskinen wrote:
>> > > Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
>> > > to be loaded to get the PHY going and ohci-omap should wait for those.
>> > > 
>> > > Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
>> > 
>> > Is this a new bug?  The 770 has been around for forever, why has this
>> > not been a problem before now?
>> 
>> PHY/OTG support for 770 has been around only since v3.14.
>
> 3.14 was released March, 2014.  A long time ago in kernel development :)
>
>> I was previously compiling phy-tahvo as built-in, and bumped into this
>> issue when compiling it as a module.
>
> Why not just make the defconfig as built-in?
>
> I'm not objecting to this patch, just really curious why no one else has
> ever hit it.

Aaro is likely to be the last active OMAP1 user (let alone hacker)
around. This is probably just a reflection of that fact. No-one else was
using N770 with upstream (at least with USB in mind) :-)
Aaro Koskinen Jan. 3, 2017, 8:12 p.m. UTC | #5
Hi,

On Tue, Jan 03, 2017 at 08:05:04PM +0100, Greg Kroah-Hartman wrote:
> On Tue, Jan 03, 2017 at 08:50:31PM +0200, Aaro Koskinen wrote:
> > On Tue, Jan 03, 2017 at 05:51:43PM +0100, Greg Kroah-Hartman wrote:
> > > On Mon, Jan 02, 2017 at 10:53:55PM +0200, Aaro Koskinen wrote:
> > > > Defer probe if PHY is missing. E.g. on Nokia 770 several modules needs
> > > > to be loaded to get the PHY going and ohci-omap should wait for those.
> > > > 
> > > > Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> > > 
> > > Is this a new bug?  The 770 has been around for forever, why has this
> > > not been a problem before now?
> > 
> > PHY/OTG support for 770 has been around only since v3.14.
> 
> 3.14 was released March, 2014.  A long time ago in kernel development :)
> 
> > I was previously compiling phy-tahvo as built-in, and bumped into this
> > issue when compiling it as a module.
> 
> Why not just make the defconfig as built-in?

I have reverted back to using built-in in my .config.

> I'm not objecting to this patch, just really curious why no one else has
> ever hit it.

Maybe I'm the only user...

The only other board using OHCI OMAP with PHY driver is H2/H3, and they
have solved the issue making USB_OHCI_HCD_OMAP1 depend on the PHY driver:

	depends on ISP1301_OMAP || !(MACH_OMAP_H2 || MACH_OMAP_H3)

This way modprobe will handle the loading order automatically.

BTW, I just realized the same issue is with omap_udc. So I guess this patch
should be just dropped, or resent so that both drivers are fixed.

A.
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Tony Lindgren Jan. 4, 2017, 5:41 p.m. UTC | #6
* Aaro Koskinen <aaro.koskinen@iki.fi> [170103 12:13]:
> On Tue, Jan 03, 2017 at 08:05:04PM +0100, Greg Kroah-Hartman wrote:
> > I'm not objecting to this patch, just really curious why no one else has
> > ever hit it.
> 
> Maybe I'm the only user...

I still have 770 in my test rack :) I don't boot it much though as
it needs a smaller kernel than what I typically build..

Tony
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/usb/host/ohci-omap.c b/drivers/usb/host/ohci-omap.c
index 495c145..fe6f6b5 100644
--- a/drivers/usb/host/ohci-omap.c
+++ b/drivers/usb/host/ohci-omap.c
@@ -230,8 +230,7 @@  static int ohci_omap_reset(struct usb_hcd *hcd)
 				return status;
 			}
 		} else {
-			dev_err(hcd->self.controller, "can't find phy\n");
-			return -ENODEV;
+			return -EPROBE_DEFER;
 		}
 		ohci->start_hnp = start_hnp;
 	}